What is generative engine optimization (GEO)?
Generative engine optimization is making content that AI search engines quote and cite. Where the term comes from, how it relates to AEO, and how to start.
What is generative engine optimization?
At a glance
| Origin | GEO: Generative Engine Optimization, arXiv, November 2023 |
|---|---|
| Same as | Answer engine optimization (AEO) |
| Engines | ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews and AI Mode |
| Manifold tools | aeo_run_ai_answers, aeo_get_site_readiness |
Where the term comes from
The name comes from a research paper, GEO: Generative Engine Optimization, first posted in November 2023 and accepted at KDD 2024. It treats AI search engines that write an answer from several sources as "generative engines", and tests which changes to a source's content make it more visible in their answers.
GEO and AEO
Marketers use GEO and AEO, answer engine optimization, for the same work: getting your brand named and your pages cited in AI answers. Pick one term for your team and use it; the method does not change. See answer engine optimization for how to measure it.
What GEO work looks like
- Check the engines can read you. AI crawlers allowed in
robots.txtand at your CDN, content in the HTML without JavaScript. - Answer the question in the first lines. Engines quote passages that answer the question directly, so put the answer before the background.
- Be in the sources the engines cite. For your buyers' questions, find the pages the engines link and get listed, mentioned or answered there.
- Measure on a fixed set of questions. Mention rate and citation rate, re-run every few weeks.
Doing it with Manifold
aeo_get_site_readiness checks the first point for free. aeo_run_ai_answers runs your buyers' questions on six engines and returns the brands named and the pages cited: it measures the last point and lists the sources to work on for the third. Getting listed on those sources is work your agent can plan and you do.
